- Email: [email protected]
CHOICE OF CONSERVATIVE OR SURGICAL TREATMENT IN BLEEDING PEPTIC ULCER
1292 the result of an analysis of the treated conservatively in Copenhagen. The series included 31 fatal cases, and 90-100% of the patients who, in my opinion, might have been aaved by appropriate surgery did not die until one or several brisk hæmatemeses had occurred in hospital. Fatal cases showing melscna alone are rare (Pedersen 1941, 1949) and generally inoperable. In the present investigation I studied the prognostic significance of factors which had not been studied in my previous analysis (Pedersen 1949) to see whether the criteria for operation then suggested needed any alteration.
This
CHOICE OF CONSERVATIVE OR SURGICAL TREATMENT IN BLEEDING PEPTIC ULCER *
JØRGFEN PEDERSEN M.B. Copenhagen ASSISTANT
SENIOR
PHYSICIAN, MEDICAL DEPARTMENT HOSPITAL, COPENHAGEN
E,
FREDERIKSBERG
THE mortality attending gastric surgery has much decreased of late, and in a few hospitals the treatment of acute bleeding ulcer has therefore been modified, consisting now of surgical as well as conservative measures. The great majority of cases are still treated conservatively in the acute phase ; surgery is reserved for the relatively few in which the prognosis with medical treatment is regarded as poor. The introduction of this surgical treatment has led to renewed interest in the prognosis of bleeding peptic ulcer. In a small group of patients who are elderly or have persisted or repeated bleeding the mortality is normally high (about 30%). Such patients are therefore potential subjects for surgery ; but the final selection for operation is made according to more definite criteria. The time of operation is a key problem ; for the operation mortality depends on how often or how long the patient has been in a state of shock. It is important to detect at an early stage of the bleeding those cases which require operation, and to operate at once ; and it is equally important not to operate in the acute phase on patients who will recover with conservative treatment. The surgical treatment can be divided into two types : surgery after trial, and immediate surgery. " Surgery after trial " means that conservative treatment (bloodtransfusion) has been tried to stop the bleeding, and that no operation is done unless conservative treatment has been unsuccessful. " Immediate surgery " means that the patient undergoes operation as soon as possible after admission to hospital with massive hsemorrhage ; here the amount of bleeding is the decisive factor. Surgery after trial is advocated by Avery Jones (1947) and
proposal
was
mortality of bleeding peptic ulcer
MATERIAL
The material consisted of the same series as beforei.e., patients in medical department B (Prof. E. Meulen. gracht) of Bispebjerg Hospital, Copenhagen, from May 1, 1938, to May 1, 1948. It includes all the patients aged 40 or more who had manifest bleeding of at least 1 litre either as hæmatemesis or as mel2ena and showed signs of peptic ulcer either on radiography or at necropsy. The group comprises 317 admissions for haemorrhage in a total of 288 patients. During the period under discussion 31 patients died in connection with bleeding TABLE II-ANALYSIS OF FACTORS
(LISTED
IN TABLE
I)
WHICH
INCREASED THE MORTALITY
peptic ulcer (found by the perusal of about 2500 necropsy records) and no less than 30 of these 31 patients were in the group selected for analysis ; 1 was aged less than 40 and thus is not. included in the series. TREATMENT
TABLE
I-PROGNOSTIC
SIGNIFICANCE
OF
VARIOUS
SYMPTOMS
AND SIGNS
The treatment
was
conservative, according to Meulen-
gracht’s (1947) principle of early and liberal feeding, administration of ample fluid, and blood-transfusion. Transfusions were given more often during the latter part of the period than at the outset. Only 1 patient was operated on. ANALYSIS
analysis comprises only 315 haemorrhages (28 2 patients died, from severe meningitis and from severe heart-disease, without the haemorrhage being manifest in life and so are excluded. The following factors were studied : (1) perforations ; (2) haemorrhage starting in hospital ; (3) severe pain in hospital; (4) stenosis and/or cachexia ; (5) operation for peptic ulcer; and (6) manifest haemorrhage in the past (tables i-v). In table I, if a patient has pyloric stenosis as well as pain he is included in both groups. The two perforations The
fatal) ;
*
Dunphy Finsterer
Data
lacking
in 1 fatal
case.
and Hoerr (1948), and immediate surgery by (1947), Tanner (1949a), and Stewart et al.
(1948).
Supported by
were
chronic, and there
was
a
grant from Kong Christian den Tiendes Fond.
no
acute diffuse
peritonitis. " Bleeding starting in hospital " includes bleeding during ulcer management, interval bleeding," and bleeding starting in another department (7patients were transferred from a surgical department, and 1 from a psychiatric department). Patients who have had With multiple operations are listed only once. partial gastrectomy taking precedence over gastroenterostomy, the latter over suture, and suture over other minor interventions, of the 44 patients listed 5 had had partial gastrectomy, 29 gastro-enterostomy, 7 simple suture, and 3 some other operation. "
have previously proposed (Pedersen 1949) the following criteria for operation in cases of acute bleeding peptic ulcer : (1) patients aged 50 or more, with a definite peptic ulcer, should be submitted to operationin the absence of serious centra-indications—if haematemesis occurs in hospital after admission ; and (2) operation should be done as soon as possible after the first hæmatemesis in hospital. I
*
listed
1293
,
TABLE III-MORTALITY IN CASES WHERE BLEEDING RECURS TO HOSPITAL OR THERE OR STARTS AFTER ADMISSION ARE OTHER FACTORS INCREASING THE MORTALITY ; AND THE SIGNIFICANCE OF THE TYPE OF BLEEDING
Nearly all the operations had
been done
a
considerable
time before.
Table 11 shows the patients listed under only one symptom, even though they had others; and the symptoms take precedence in the order shown from top to bottom-e.g., a patient with pain and cachexia is listed only under pain. The age-incidence is not shown but is of the usual significance. -
The first four factors listed in table i increased the and must be taken to mean that the patient had a severe, penetrating, and usually chronic ulcer. This accords with the findings of other workers (Allen and Benedict 1933, Baker 1947, Finsterer 1947, GordonTaylor 1946, Avery Jones 1947). In the present series the factors increasing the mortality may be reduced to severe pain and/or haemorrhage starting in hospital (table II) ; in the presence of one or both of these factors the mortality from hsematemesis after admission to hospital was about 60% (tables in and IV). In hospital the incidence of these factors in the presence of melsena is comparatively higher than in haematemesis, yet the mortality is considerably lower (table III); but the patients with melaena were on the average somewhat younger than those with The group of 68 patients who had haematemesis. hæmatemesis in hospital, with a mortality of 31% (table IV), is divisible, by other easily recognisable signs (haemorrhage starting in hospital and/or severe pain), into 16 with a mortality of 63% and 52 with a mortality
mortality,
21% (table iv). 14% had previously undergone operation for peptic ulcer, and the gross mortality among them was the of
same as that of the entire series. Little interest seems to have been shown in this aspect. Welch and Yunich (1940) had no deaths among their patients with bleeding peptic ulcer who had previously undergone gastro-
enterostomy. TABLE
IV-MORTALITY
FROM
HÆMATEMESIS
IN
COMBINED WITH OTHER SYMPTOMS
*
Data lacking in the
case
of
one
death.
HOSPITAL
The incidence of previous attacks of bleeding is very as Rischel (1936) and Hesser (1942) also found in their series. Among the " relapsers " the mortality is lower, if anything, than among the patients bleeding for the first time. Of 27 deaths 17 took place in connection with the first episode of haemorrhage. A similar finding has been reported by Christiansen (1935), Allen and Benedict (1933), Blackford and Allen (1942), Welch (1949), and Baker (1947), but not by Sullens et al. (1949). It is clear from table v that the prognosis is the same, whether manifest bleeding has taken place once or three times or more, if only the present bleeding is in the form of hsematemesis. Every time a haemorrhage takes the has the same chance—i.e., that patient place, appropriate to his age-group. It will be seen from the table that the difference in the mortality among patients admitted to hospital during the first attack (17 out of 164) and those who have had one or several attacks before (10 out of 150), is due almost exclusively to a difference in mortality from haematemesis and from melaena alone.
high,
Conclusion The mortality is increased in the presence of the
following symptoms : perforation, severe pain in hospital, haemorrhage starting in hospital, stenosis, and cachexia. No such increase is caused by previous operations for peptic ulcer (gastro-enterostomy) or previous manifest bleeding. The combination of haematemesis in hospital with severe pain in hospital or with haemorrhage starting in hospital entailed a mortality of 60%. TABLE
V-MORTALITY-RATE IN RELATION ATTACKS OF HEMORRHAGE
* Data not available in
one
excluded
TO
PREVIOUS
case.
This
analysis illustrates how impossible it is to compare mortality statistics, especially gross mortalities (Avery Jones 1947, Pedersen 1941, 1949).
most
-
CRITERIA OF OPERATION
Not all the results
obtained by the present analysis applicable directly to the criteria of operation. The present analysis was based on the entire series, including patients who did not die from exsanguination and nonoperable patients. The criteria of operation, on the other hand, can be based only on operable patients with haemorrhage after admission and, according to my previous investigations (Pedersen 1941, 1949), only on operable patients with haematemesis in hospital. This reduces the material which can be used in assessing the criteria of operation from 68 with hsematemesis in hospital of whom 21 died (table Iv) to 50 of whom 9 died. (For my estimation of the operability among the are
deceased and the survivors see Pedersen 1949.) As might be expected, the significance of the factors studied is restricted by their occurrence in relatively few of the deceased patients who had been operable. Therefore, the entire analysis will not be repeated on the basis of the relevant material, but table vi illustrates the result of a study of a few factors in this material.
1294 OF
SYMPTOMS INCREASING TREATMENT OF OPERABLE CASES OF BLEEDING PEPTIC ULCER
TABLE
VI-SIGNIFICANCE
THE MORTALITY-RATE
FOR
SOME
CONSERVATIVE
the time chosen for operation. In Tanner’s (1949b) opinion, at immediate surgery 65-85% of the diagnoses are correct. In the criteria previously proposed the age-limit was based on the mortality with conservative treatment of patients who can be made operable after one hæmatemesis in hospital as’ compared with the presumed mortality from routine partial gastrectomies at centres for gastric surgery. In the age-group above 50 I found (Pedersen 1949) a mortality of 25% in conservative treatment and set 10% as the maximum limit of the routine operative mortality, if surgery was to be embarked on at all. Since a Danish report has now been published of the operative mortality from partial gastrectomy (Køster 1951), a comparison is appropriate. Such comparison, illustrated in table vn, shows that for the present there is no basis for proposing " routine " surgery in bleeding peptic ulcer in patients aged 70 or more, although they need operation more than others do. The ordinary mortality attending gastric surgery in this -age-group is not yet sufficiently known, because the numbers ’. involved are too small. The surgical statistics in table vn give a too optimistic picture of the results in cases of haemorrhage, and-the conservative statistics. are not ideal (final diagnoses). Besides, the latter cover a period of ten years, preceding that. of the surgical statistics, and in the course of this period the treatment was altered in one respectbloodtransfusions were, adopted in increasing numbers. Some of the patients have received too little blood, judging by our present standards. It is impossible to decide the exact importance of this factor, but undoubtedly a number of patients with bleeding peptic ulcer die despite the most heroic transfusions. ’
,
It will be seen from table vi that operation is indicated in those aged 40-49, if- there is shock and pain in addition to haematemesis, but hardly in the presence of shock alone. In patients aged 50 or more it is not advisable to follow the principle of awaiting a second hæmatemesis in hospital ; I have seen 1 case ill which the patient .died from the second hæmatemesis, and I have heard of 2—i.e., 3 cases from two hospitals in Copenhagen in eighteen months.Apparently, in the age-group 50 or more, little is to be gained by insisting that the hæmatemesis should be accompanied by haemorrhagic shock before operation is done (" entire series " in table vi as compared with " at least onehæmorrhagic shock "). If the first haematemesis entails shock, there is no doubt ; "but, if it does not, should one await a second hsematemesis which possibly leads to death ? I do not think so. If the diagnosis is considered fairly definite, operation is called for as soon as possible after the first hæmatemesis in
hospital. THE SHORTCOMINGS OF THE ANALYSIS
My series includes only definite peptic ulcers as established particularly by radiography. The diagnoses are final and therefore the calculation of the mortality and the analysis of prognostic factors in the whole series are very reliable. In deciding on operation criteria, however, one must avail oneself of initial diagnoses as emphasised inter alios by Lewin and Truelove (1949). So it has to be borne in mind that by reason of the selection of the material no regard has been paid to the difficulties of establishing a diagnosis of chronic ulcer during an acute bleeding. If we use. surgery some superficial acute ulcers will wrongly be operated on ; on the other hand, there is some danger that chronic ulcers may not be diagnosed and operated on early enough. The errors in diagnosis depend somewhat on TABLE VII-MORTALITY OF BLEEDING PEPTIC ULCER TREATED ,
,
CONSERVATIVELY AND OF PARTIAL GASTRECTOMY FOR PEPTIC ULCER ’
.
AMENDED .CRITERIA In the light of what has been said above, I feel that The criteria my criteria should be altered slightly. presuppose : every probability of chronic peptic ulcer, ample transfusions immediately after admission, and no serious contra-indications to surgery.In the age-group 50-7.0 surgery is called for immediately after the first hsematemesis in hospital. This applies unconditionally in all cases where the hæmatemesis entails haemorrhagic shock or has started in hospital, or where the patient has severe pain in hospital or has pyloric stenosis. It is recommended in other cases of haematemesis in hospital, when chronic peptic ulcer is merely considered very probable. Patients aged 40-50 should be operated on immediately after the first haematemesis only if several of the absolute conditions are fulfilled. In the age-group below 40, operation is considered only.in exceptional cases and only after lengthy attempts at conservative treatment.. For patients aged 70 or more operation cannot be recommended as a routine at present, because the series studied so far are so small that they allow only the conclusion that the mortality is very high both in conservative treatment of operable patients with haematemesis in hospital and in operation for peptic ulcer generally. If surgery -is to be tried in this age-group, it should be done immediately after the first haematemsis in hospital, as in the age-group ’
50-70.
(1951).
.
,
These criteria approach those advanced by Romcke and Avery Jones (1947) ; both emphasise the importance of debilitating recurrent haemorrhages, but they do not say at what -stage the patients should be operated on. The mode of manifestation of the recurrent haemorrhage makes no real difference-Avery Jones (personal communication) found that of 28 patients operated on 25 had had hæmatemesis in hospital. In Avery Jones’s (1947) series the patients underwent operation after having had 1-4 (average 2-4) brisk bleedings after admission. Avery Jones estimates a
(1945)
* After Køster
’
1295
mortality
-
of
20-30%
for the rather late
operations
and
restricts the use of surgery, whereas I suggest a somewhat extended and earlier use of surgery and therefore may estimate a lower mortality (about 10% for operation after one haemorrhage in hospital). Amendola (1949) and Welch (1949) suggest criteria which accord well with those set out here. Both regard recurrent bleeding as an indication for immediate
patients aged more than 50. severe exsanguination on admission and where there is difficulty in keeping bleeding persistent the patient out of haemorrhagic shock, operation within 24-48 hours of admission is recommended by several workers, if the patient cannot be stabilised by at least
operation In
in
cases
of
2500 ml. of blood. Haemorrhages of this kind are rare in Copenhagen, and no such deaths occurred in the present series. CONCLUSION
Whether a combination of conservative measures and surgery affords better results than a purely conservative treatment is a question which cannot yet be answered, and the results published do not offer much help. The object is -not to obtain a low operation mortality, but to obtain a lower mortality by choosing between surgical and conservative treatment than by conservative treatment alone. This depends on selecting the right patients for operation, and it might perhaps be compatible with a
comparatively high operation mortality. For the time being, a combination of conservative measures and surgery must be considered a justified therapeutic experiment. The operations are difficult and few, and should be undertaken only by surgeons with a, wide experience in gastric surgery in departments fitted with the best equipment and in collaboration with physicians. To assess the " combined " therapy it is necessary also to have departments serving the same population and offering the same conditions of stay in hospital and facilities (blood-bank) over. the same period. A comparison of conservative and conservative-surgical treatment cannot be made retrospectively, inter aliabecause the age of those who die from bleeding peptic ulcer is becoming higher all -the time, and because the’ treatment is rapidly changing. For some time surgery in bleeding peptic ulcer has been given a trial in,Copenhagen according to the indications proposed, aridthe results will be published by Køster (see Køster 1950). ’
.
’.
’
SUMMARY
by conservative measures (Meulengracht’s régime) and, as a result, proposed criteria for surgery in the acute phase with- " haematemesis in- hospital " as the leading sign. The analysis is extended here to include factors (in addition to recurrent bleeding,. advancedage, and chronic ulcer) which increase the mortality. In the entire series the mortality is, increased in the presence- of the following symptoms : perforation, haemorrhage starting in hospital, severe pain in hospital, pyloric stenosis, and cachexia. The mortality is not increased merely because the patient has undergone operation for peptic ulcer or has had manifest bleeding in the past. The combination of haematemesis in hospital and severe pain in hospital was attended by a mortality of 60%. A proposed modification of the criteria for operation presupposes a good clinical probability of chronic peptic ulcer, ample transfusions immediately after admission, treated
no
criteria
serious
contra-indications
to
surgery. - The
follow : in the age-group 50-70 operation is called for immediately after the first haematemesis are
ui hospital.
Allen, A. W., Benedict, E. B. (1933) Ann. Surg. 98, 736. Amendola, F. H. (1949) Ibid, 129, 47. Baker, C. (1947) Guy’s Hosp. Rep. 96, 1. Blackford, J. M., Allen, H. E. (1942) J. Amer. med. Ass. 120, 811. Christiansen, T. (1935) Acta med. scand. 84, 374. Dunphy, J. E., Hoerr, S. O. (1948) Surgery, 24, 231. Finsterer, H. (1947) Wien. med. Wschr. 97, 3. Gordon-Taylor, G. (1946) Brit. J. Surg. 33, 336. Hesser, S. (1942) Svenska Läkartidn. 39, 141. Jones, A. (1947) Brit. med. J. ii, 441, 477. Køster, K. H. (1950) Lancet, i, 831. (1951) Ugeskr. Laeg. 113, 144. Lewin, D. C., Truelove, S. (1949) Brit. med. J. i, 383. Meulengracht, E. (1947) Arch. intern. Med. 80, 697. Pedersen, J. (1941) Nord. Med. 9, 252. (1949) Gastroenterology, 12, 597. Rischel, A. (1936) Klin. Wschr. 15, 335. Römcke, O. (1945) Nord. Med. 27, 1533. Stewart, J. D., Schaer, S. M., Potter, W. H., Massover, A. J. (1948) Ann. Surg. 128, 791. Sullens, W. E., Steigmann, F., Meyer, K. A. (1949) Arch. Surg. 59, -
-
1244.
Tanner, N. C. (1949a) Med. Pr. 221, 9. (1949b) Lancet, ii, 1135. Welch, C. E. (1949) J. Amer. med. Ass. 141, 1113. Yunich, A. M. (1940) Surg. Gynec. Obstet. 70, 662. -
—
METABOLIC DISORDERS IN HEAD INJURY HYPERCHLORÆMIA AND HYPOCHLORURIA
G. HIGGINS
WALPOLE LEWIN
B.Sc.
M.S. Lond., F.R.C.S.
W.
J. R. P. O’BRIEN
in all
cases
H. TAYLOR
M.A., B.M. Oxfd, M.R.C.P.
M.A., B.Sc.
Department
of Biochemistry,
Radcliffe Infirmary, Oxford METABOLIC disorders may accompany coma from whatever cause, and after head injury they-are most likely to occur in those patients who survive the. early hours after the accident yet remain unconscious. Bio. chemical studies have been made in such -patients admitted to the accident service of this hospital and thispaper records an unusual complication found in some of them, characterised by the presence of a high plasma chloride-ion ’concentration in the absence of any: significant excretion of the ion in the urine. It is difficult to assess the significance of these changes in such injured patients, some of whom in addition to the head injury suffered from shock and other injuries or from chest complications ; nor is it possible from the small series so far studied to determine the frequency of these disturbances. Nevertheless, these cases are reported at this stage to direct further attention to these problems and to discuss the diagnostic and
severely
’
therapeutic implications. Case-reports
as
This applies., unconditionally
’
REFERENCES
From the Accident Service and
In a previous paper (Pedersen 1949) I analysed the mortality among patients with bleeding peptic ulcer
and
where haematemesis is accompanied by haemorrhagic shock, or if the patient has severe pain in hospital or pyloric stenosis. It is, moreover, recommended in other cases of haematemesis in hospital when the diagnosis of chronic peptic ulcer is regarded as certain or highly probable. In the age-group 40-50 operation should be done immediately after the first hæmatemesis only if the patient fulfils several of the absolute criteria. For patients aged less than 40 operation is considered only in exceptional cases and only after lengthy conservative treatment. For patients aged 70 or more, operation cannot be recommended as a routine because the series hitherto studied are so small that they lead only to the conclusion that the mortality is very high both in conservative treatment of operable patients with haematemesis in hospital and in operations for peptic ulcer generally. If surgery is to be tried in this age-group, the operation should be done immediately after the first hæmatemesis in hospital, as in the age-group 50-70. Surgery in bleeding peptic ulcer is under trial along these lines (Køster 1950, 1951).
Six
cases are
- output
reported.
were recorded
as
The fluid intake and
accurately
as
urinary
circumstances
This
CHOICE OF CONSERVATIVE OR SURGICAL TREATMENT IN BLEEDING PEPTIC ULCER *
JØRGFEN PEDERSEN M.B. Copenhagen ASSISTANT
SENIOR
PHYSICIAN, MEDICAL DEPARTMENT HOSPITAL, COPENHAGEN
E,
FREDERIKSBERG
THE mortality attending gastric surgery has much decreased of late, and in a few hospitals the treatment of acute bleeding ulcer has therefore been modified, consisting now of surgical as well as conservative measures. The great majority of cases are still treated conservatively in the acute phase ; surgery is reserved for the relatively few in which the prognosis with medical treatment is regarded as poor. The introduction of this surgical treatment has led to renewed interest in the prognosis of bleeding peptic ulcer. In a small group of patients who are elderly or have persisted or repeated bleeding the mortality is normally high (about 30%). Such patients are therefore potential subjects for surgery ; but the final selection for operation is made according to more definite criteria. The time of operation is a key problem ; for the operation mortality depends on how often or how long the patient has been in a state of shock. It is important to detect at an early stage of the bleeding those cases which require operation, and to operate at once ; and it is equally important not to operate in the acute phase on patients who will recover with conservative treatment. The surgical treatment can be divided into two types : surgery after trial, and immediate surgery. " Surgery after trial " means that conservative treatment (bloodtransfusion) has been tried to stop the bleeding, and that no operation is done unless conservative treatment has been unsuccessful. " Immediate surgery " means that the patient undergoes operation as soon as possible after admission to hospital with massive hsemorrhage ; here the amount of bleeding is the decisive factor. Surgery after trial is advocated by Avery Jones (1947) and
proposal
was
mortality of bleeding peptic ulcer
MATERIAL
The material consisted of the same series as beforei.e., patients in medical department B (Prof. E. Meulen. gracht) of Bispebjerg Hospital, Copenhagen, from May 1, 1938, to May 1, 1948. It includes all the patients aged 40 or more who had manifest bleeding of at least 1 litre either as hæmatemesis or as mel2ena and showed signs of peptic ulcer either on radiography or at necropsy. The group comprises 317 admissions for haemorrhage in a total of 288 patients. During the period under discussion 31 patients died in connection with bleeding TABLE II-ANALYSIS OF FACTORS
(LISTED
IN TABLE
I)
WHICH
INCREASED THE MORTALITY
peptic ulcer (found by the perusal of about 2500 necropsy records) and no less than 30 of these 31 patients were in the group selected for analysis ; 1 was aged less than 40 and thus is not. included in the series. TREATMENT
TABLE
I-PROGNOSTIC
SIGNIFICANCE
OF
VARIOUS
SYMPTOMS
AND SIGNS
The treatment
was
conservative, according to Meulen-
gracht’s (1947) principle of early and liberal feeding, administration of ample fluid, and blood-transfusion. Transfusions were given more often during the latter part of the period than at the outset. Only 1 patient was operated on. ANALYSIS
analysis comprises only 315 haemorrhages (28 2 patients died, from severe meningitis and from severe heart-disease, without the haemorrhage being manifest in life and so are excluded. The following factors were studied : (1) perforations ; (2) haemorrhage starting in hospital ; (3) severe pain in hospital; (4) stenosis and/or cachexia ; (5) operation for peptic ulcer; and (6) manifest haemorrhage in the past (tables i-v). In table I, if a patient has pyloric stenosis as well as pain he is included in both groups. The two perforations The
fatal) ;
*
Dunphy Finsterer
Data
lacking
in 1 fatal
case.
and Hoerr (1948), and immediate surgery by (1947), Tanner (1949a), and Stewart et al.
(1948).
Supported by
were
chronic, and there
was
a
grant from Kong Christian den Tiendes Fond.
no
acute diffuse
peritonitis. " Bleeding starting in hospital " includes bleeding during ulcer management, interval bleeding," and bleeding starting in another department (7patients were transferred from a surgical department, and 1 from a psychiatric department). Patients who have had With multiple operations are listed only once. partial gastrectomy taking precedence over gastroenterostomy, the latter over suture, and suture over other minor interventions, of the 44 patients listed 5 had had partial gastrectomy, 29 gastro-enterostomy, 7 simple suture, and 3 some other operation. "
have previously proposed (Pedersen 1949) the following criteria for operation in cases of acute bleeding peptic ulcer : (1) patients aged 50 or more, with a definite peptic ulcer, should be submitted to operationin the absence of serious centra-indications—if haematemesis occurs in hospital after admission ; and (2) operation should be done as soon as possible after the first hæmatemesis in hospital. I
*
listed
1293
,
TABLE III-MORTALITY IN CASES WHERE BLEEDING RECURS TO HOSPITAL OR THERE OR STARTS AFTER ADMISSION ARE OTHER FACTORS INCREASING THE MORTALITY ; AND THE SIGNIFICANCE OF THE TYPE OF BLEEDING
Nearly all the operations had
been done
a
considerable
time before.
Table 11 shows the patients listed under only one symptom, even though they had others; and the symptoms take precedence in the order shown from top to bottom-e.g., a patient with pain and cachexia is listed only under pain. The age-incidence is not shown but is of the usual significance. -
The first four factors listed in table i increased the and must be taken to mean that the patient had a severe, penetrating, and usually chronic ulcer. This accords with the findings of other workers (Allen and Benedict 1933, Baker 1947, Finsterer 1947, GordonTaylor 1946, Avery Jones 1947). In the present series the factors increasing the mortality may be reduced to severe pain and/or haemorrhage starting in hospital (table II) ; in the presence of one or both of these factors the mortality from hsematemesis after admission to hospital was about 60% (tables in and IV). In hospital the incidence of these factors in the presence of melsena is comparatively higher than in haematemesis, yet the mortality is considerably lower (table III); but the patients with melaena were on the average somewhat younger than those with The group of 68 patients who had haematemesis. hæmatemesis in hospital, with a mortality of 31% (table IV), is divisible, by other easily recognisable signs (haemorrhage starting in hospital and/or severe pain), into 16 with a mortality of 63% and 52 with a mortality
mortality,
21% (table iv). 14% had previously undergone operation for peptic ulcer, and the gross mortality among them was the of
same as that of the entire series. Little interest seems to have been shown in this aspect. Welch and Yunich (1940) had no deaths among their patients with bleeding peptic ulcer who had previously undergone gastro-
enterostomy. TABLE
IV-MORTALITY
FROM
HÆMATEMESIS
IN
COMBINED WITH OTHER SYMPTOMS
*
Data lacking in the
case
of
one
death.
HOSPITAL
The incidence of previous attacks of bleeding is very as Rischel (1936) and Hesser (1942) also found in their series. Among the " relapsers " the mortality is lower, if anything, than among the patients bleeding for the first time. Of 27 deaths 17 took place in connection with the first episode of haemorrhage. A similar finding has been reported by Christiansen (1935), Allen and Benedict (1933), Blackford and Allen (1942), Welch (1949), and Baker (1947), but not by Sullens et al. (1949). It is clear from table v that the prognosis is the same, whether manifest bleeding has taken place once or three times or more, if only the present bleeding is in the form of hsematemesis. Every time a haemorrhage takes the has the same chance—i.e., that patient place, appropriate to his age-group. It will be seen from the table that the difference in the mortality among patients admitted to hospital during the first attack (17 out of 164) and those who have had one or several attacks before (10 out of 150), is due almost exclusively to a difference in mortality from haematemesis and from melaena alone.
high,
Conclusion The mortality is increased in the presence of the
following symptoms : perforation, severe pain in hospital, haemorrhage starting in hospital, stenosis, and cachexia. No such increase is caused by previous operations for peptic ulcer (gastro-enterostomy) or previous manifest bleeding. The combination of haematemesis in hospital with severe pain in hospital or with haemorrhage starting in hospital entailed a mortality of 60%. TABLE
V-MORTALITY-RATE IN RELATION ATTACKS OF HEMORRHAGE
* Data not available in
one
excluded
TO
PREVIOUS
case.
This
analysis illustrates how impossible it is to compare mortality statistics, especially gross mortalities (Avery Jones 1947, Pedersen 1941, 1949).
most
-
CRITERIA OF OPERATION
Not all the results
obtained by the present analysis applicable directly to the criteria of operation. The present analysis was based on the entire series, including patients who did not die from exsanguination and nonoperable patients. The criteria of operation, on the other hand, can be based only on operable patients with haemorrhage after admission and, according to my previous investigations (Pedersen 1941, 1949), only on operable patients with haematemesis in hospital. This reduces the material which can be used in assessing the criteria of operation from 68 with hsematemesis in hospital of whom 21 died (table Iv) to 50 of whom 9 died. (For my estimation of the operability among the are
deceased and the survivors see Pedersen 1949.) As might be expected, the significance of the factors studied is restricted by their occurrence in relatively few of the deceased patients who had been operable. Therefore, the entire analysis will not be repeated on the basis of the relevant material, but table vi illustrates the result of a study of a few factors in this material.
1294 OF
SYMPTOMS INCREASING TREATMENT OF OPERABLE CASES OF BLEEDING PEPTIC ULCER
TABLE
VI-SIGNIFICANCE
THE MORTALITY-RATE
FOR
SOME
CONSERVATIVE
the time chosen for operation. In Tanner’s (1949b) opinion, at immediate surgery 65-85% of the diagnoses are correct. In the criteria previously proposed the age-limit was based on the mortality with conservative treatment of patients who can be made operable after one hæmatemesis in hospital as’ compared with the presumed mortality from routine partial gastrectomies at centres for gastric surgery. In the age-group above 50 I found (Pedersen 1949) a mortality of 25% in conservative treatment and set 10% as the maximum limit of the routine operative mortality, if surgery was to be embarked on at all. Since a Danish report has now been published of the operative mortality from partial gastrectomy (Køster 1951), a comparison is appropriate. Such comparison, illustrated in table vn, shows that for the present there is no basis for proposing " routine " surgery in bleeding peptic ulcer in patients aged 70 or more, although they need operation more than others do. The ordinary mortality attending gastric surgery in this -age-group is not yet sufficiently known, because the numbers ’. involved are too small. The surgical statistics in table vn give a too optimistic picture of the results in cases of haemorrhage, and-the conservative statistics. are not ideal (final diagnoses). Besides, the latter cover a period of ten years, preceding that. of the surgical statistics, and in the course of this period the treatment was altered in one respectbloodtransfusions were, adopted in increasing numbers. Some of the patients have received too little blood, judging by our present standards. It is impossible to decide the exact importance of this factor, but undoubtedly a number of patients with bleeding peptic ulcer die despite the most heroic transfusions. ’
,
It will be seen from table vi that operation is indicated in those aged 40-49, if- there is shock and pain in addition to haematemesis, but hardly in the presence of shock alone. In patients aged 50 or more it is not advisable to follow the principle of awaiting a second hæmatemesis in hospital ; I have seen 1 case ill which the patient .died from the second hæmatemesis, and I have heard of 2—i.e., 3 cases from two hospitals in Copenhagen in eighteen months.Apparently, in the age-group 50 or more, little is to be gained by insisting that the hæmatemesis should be accompanied by haemorrhagic shock before operation is done (" entire series " in table vi as compared with " at least onehæmorrhagic shock "). If the first haematemesis entails shock, there is no doubt ; "but, if it does not, should one await a second hsematemesis which possibly leads to death ? I do not think so. If the diagnosis is considered fairly definite, operation is called for as soon as possible after the first hæmatemesis in
hospital. THE SHORTCOMINGS OF THE ANALYSIS
My series includes only definite peptic ulcers as established particularly by radiography. The diagnoses are final and therefore the calculation of the mortality and the analysis of prognostic factors in the whole series are very reliable. In deciding on operation criteria, however, one must avail oneself of initial diagnoses as emphasised inter alios by Lewin and Truelove (1949). So it has to be borne in mind that by reason of the selection of the material no regard has been paid to the difficulties of establishing a diagnosis of chronic ulcer during an acute bleeding. If we use. surgery some superficial acute ulcers will wrongly be operated on ; on the other hand, there is some danger that chronic ulcers may not be diagnosed and operated on early enough. The errors in diagnosis depend somewhat on TABLE VII-MORTALITY OF BLEEDING PEPTIC ULCER TREATED ,
,
CONSERVATIVELY AND OF PARTIAL GASTRECTOMY FOR PEPTIC ULCER ’
.
AMENDED .CRITERIA In the light of what has been said above, I feel that The criteria my criteria should be altered slightly. presuppose : every probability of chronic peptic ulcer, ample transfusions immediately after admission, and no serious contra-indications to surgery.In the age-group 50-7.0 surgery is called for immediately after the first hsematemesis in hospital. This applies unconditionally in all cases where the hæmatemesis entails haemorrhagic shock or has started in hospital, or where the patient has severe pain in hospital or has pyloric stenosis. It is recommended in other cases of haematemesis in hospital, when chronic peptic ulcer is merely considered very probable. Patients aged 40-50 should be operated on immediately after the first haematemesis only if several of the absolute conditions are fulfilled. In the age-group below 40, operation is considered only.in exceptional cases and only after lengthy attempts at conservative treatment.. For patients aged 70 or more operation cannot be recommended as a routine at present, because the series studied so far are so small that they allow only the conclusion that the mortality is very high both in conservative treatment of operable patients with haematemesis in hospital and in operation for peptic ulcer generally. If surgery -is to be tried in this age-group, it should be done immediately after the first haematemsis in hospital, as in the age-group ’
50-70.
(1951).
.
,
These criteria approach those advanced by Romcke and Avery Jones (1947) ; both emphasise the importance of debilitating recurrent haemorrhages, but they do not say at what -stage the patients should be operated on. The mode of manifestation of the recurrent haemorrhage makes no real difference-Avery Jones (personal communication) found that of 28 patients operated on 25 had had hæmatemesis in hospital. In Avery Jones’s (1947) series the patients underwent operation after having had 1-4 (average 2-4) brisk bleedings after admission. Avery Jones estimates a
(1945)
* After Køster
’
1295
mortality
-
of
20-30%
for the rather late
operations
and
restricts the use of surgery, whereas I suggest a somewhat extended and earlier use of surgery and therefore may estimate a lower mortality (about 10% for operation after one haemorrhage in hospital). Amendola (1949) and Welch (1949) suggest criteria which accord well with those set out here. Both regard recurrent bleeding as an indication for immediate
patients aged more than 50. severe exsanguination on admission and where there is difficulty in keeping bleeding persistent the patient out of haemorrhagic shock, operation within 24-48 hours of admission is recommended by several workers, if the patient cannot be stabilised by at least
operation In
in
cases
of
2500 ml. of blood. Haemorrhages of this kind are rare in Copenhagen, and no such deaths occurred in the present series. CONCLUSION
Whether a combination of conservative measures and surgery affords better results than a purely conservative treatment is a question which cannot yet be answered, and the results published do not offer much help. The object is -not to obtain a low operation mortality, but to obtain a lower mortality by choosing between surgical and conservative treatment than by conservative treatment alone. This depends on selecting the right patients for operation, and it might perhaps be compatible with a
comparatively high operation mortality. For the time being, a combination of conservative measures and surgery must be considered a justified therapeutic experiment. The operations are difficult and few, and should be undertaken only by surgeons with a, wide experience in gastric surgery in departments fitted with the best equipment and in collaboration with physicians. To assess the " combined " therapy it is necessary also to have departments serving the same population and offering the same conditions of stay in hospital and facilities (blood-bank) over. the same period. A comparison of conservative and conservative-surgical treatment cannot be made retrospectively, inter aliabecause the age of those who die from bleeding peptic ulcer is becoming higher all -the time, and because the’ treatment is rapidly changing. For some time surgery in bleeding peptic ulcer has been given a trial in,Copenhagen according to the indications proposed, aridthe results will be published by Køster (see Køster 1950). ’
.
’.
’
SUMMARY
by conservative measures (Meulengracht’s régime) and, as a result, proposed criteria for surgery in the acute phase with- " haematemesis in- hospital " as the leading sign. The analysis is extended here to include factors (in addition to recurrent bleeding,. advancedage, and chronic ulcer) which increase the mortality. In the entire series the mortality is, increased in the presence- of the following symptoms : perforation, haemorrhage starting in hospital, severe pain in hospital, pyloric stenosis, and cachexia. The mortality is not increased merely because the patient has undergone operation for peptic ulcer or has had manifest bleeding in the past. The combination of haematemesis in hospital and severe pain in hospital was attended by a mortality of 60%. A proposed modification of the criteria for operation presupposes a good clinical probability of chronic peptic ulcer, ample transfusions immediately after admission, treated
no
criteria
serious
contra-indications
to
surgery. - The
follow : in the age-group 50-70 operation is called for immediately after the first haematemesis are
ui hospital.
Allen, A. W., Benedict, E. B. (1933) Ann. Surg. 98, 736. Amendola, F. H. (1949) Ibid, 129, 47. Baker, C. (1947) Guy’s Hosp. Rep. 96, 1. Blackford, J. M., Allen, H. E. (1942) J. Amer. med. Ass. 120, 811. Christiansen, T. (1935) Acta med. scand. 84, 374. Dunphy, J. E., Hoerr, S. O. (1948) Surgery, 24, 231. Finsterer, H. (1947) Wien. med. Wschr. 97, 3. Gordon-Taylor, G. (1946) Brit. J. Surg. 33, 336. Hesser, S. (1942) Svenska Läkartidn. 39, 141. Jones, A. (1947) Brit. med. J. ii, 441, 477. Køster, K. H. (1950) Lancet, i, 831. (1951) Ugeskr. Laeg. 113, 144. Lewin, D. C., Truelove, S. (1949) Brit. med. J. i, 383. Meulengracht, E. (1947) Arch. intern. Med. 80, 697. Pedersen, J. (1941) Nord. Med. 9, 252. (1949) Gastroenterology, 12, 597. Rischel, A. (1936) Klin. Wschr. 15, 335. Römcke, O. (1945) Nord. Med. 27, 1533. Stewart, J. D., Schaer, S. M., Potter, W. H., Massover, A. J. (1948) Ann. Surg. 128, 791. Sullens, W. E., Steigmann, F., Meyer, K. A. (1949) Arch. Surg. 59, -
-
1244.
Tanner, N. C. (1949a) Med. Pr. 221, 9. (1949b) Lancet, ii, 1135. Welch, C. E. (1949) J. Amer. med. Ass. 141, 1113. Yunich, A. M. (1940) Surg. Gynec. Obstet. 70, 662. -
—
METABOLIC DISORDERS IN HEAD INJURY HYPERCHLORÆMIA AND HYPOCHLORURIA
G. HIGGINS
WALPOLE LEWIN
B.Sc.
M.S. Lond., F.R.C.S.
W.
J. R. P. O’BRIEN
in all
cases
H. TAYLOR
M.A., B.M. Oxfd, M.R.C.P.
M.A., B.Sc.
Department
of Biochemistry,
Radcliffe Infirmary, Oxford METABOLIC disorders may accompany coma from whatever cause, and after head injury they-are most likely to occur in those patients who survive the. early hours after the accident yet remain unconscious. Bio. chemical studies have been made in such -patients admitted to the accident service of this hospital and thispaper records an unusual complication found in some of them, characterised by the presence of a high plasma chloride-ion ’concentration in the absence of any: significant excretion of the ion in the urine. It is difficult to assess the significance of these changes in such injured patients, some of whom in addition to the head injury suffered from shock and other injuries or from chest complications ; nor is it possible from the small series so far studied to determine the frequency of these disturbances. Nevertheless, these cases are reported at this stage to direct further attention to these problems and to discuss the diagnostic and
severely
’
therapeutic implications. Case-reports
as
This applies., unconditionally
’
REFERENCES
From the Accident Service and
In a previous paper (Pedersen 1949) I analysed the mortality among patients with bleeding peptic ulcer
and
where haematemesis is accompanied by haemorrhagic shock, or if the patient has severe pain in hospital or pyloric stenosis. It is, moreover, recommended in other cases of haematemesis in hospital when the diagnosis of chronic peptic ulcer is regarded as certain or highly probable. In the age-group 40-50 operation should be done immediately after the first hæmatemesis only if the patient fulfils several of the absolute criteria. For patients aged less than 40 operation is considered only in exceptional cases and only after lengthy conservative treatment. For patients aged 70 or more, operation cannot be recommended as a routine because the series hitherto studied are so small that they lead only to the conclusion that the mortality is very high both in conservative treatment of operable patients with haematemesis in hospital and in operations for peptic ulcer generally. If surgery is to be tried in this age-group, the operation should be done immediately after the first hæmatemesis in hospital, as in the age-group 50-70. Surgery in bleeding peptic ulcer is under trial along these lines (Køster 1950, 1951).
Six
cases are
- output
reported.
were recorded
as
The fluid intake and
accurately
as
urinary
circumstances